面向特征增强的危险品运单深度分析方法

doi:10.16097/j.cnki.1009-6744.2025.04.028

交通运输系统工程与信息 ›› 2025, Vol. 25 ›› Issue (4): 306-316.DOI: 10.16097/j.cnki.1009-6744.2025.04.028

面向特征增强的危险品运单深度分析方法

陈邦举^1,2,3，罗义凯⁴，陈磊磊⁴，伍翰廷⁴，项昌乐^*1

1. 北京理工大学，机械与车辆学院，北京100081；2.中交信捷科技有限公司，北京100011； 3. 中国交通通信信息中心，北京100011；4.长安大学，运输工程学院，西安710064

收稿日期:2025-04-07 修回日期:2025-05-09 接受日期:2025-05-15 出版日期:2025-08-25 发布日期:2025-08-25
作者简介:陈邦举(1994—)，男，安徽庐江人，博士生。
基金资助:
国家自然科学基金 (72371035)；陕西省自然科学基础研究计划项目 (2020JM-237)。

A Deep Learning Approach for Enhancement of Hazardous Goods Waybill's Feature

CHEN Bangju^1,2,3, LUO Yikai⁴, CHEN Leilei⁴, WU Hanting⁴, XIANG Changle^*1

1. School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China; 2. China Transport InfoJet Technologies Co Ltd, Beijing 100011, China; 3. China Transport Telecommunications & Information Center, Beijing 100011, China; 4. School of Transportation Engineering, Chang'an University, Xi'an 710064, China

Received:2025-04-07 Revised:2025-05-09 Accepted:2025-05-15 Online:2025-08-25 Published:2025-08-25
Supported by:
National Natural Science Foundation of China (72371035)；Natural Science Basic Research Plan in Shaanxi Province (2020JM-237)。

摘要/Abstract

摘要： 我国危险货物运输的运单常由运输企业在运输前填写报备，其起讫点位置及时间不精确，在运输管理方面发挥作用有限。为提升危险品运单质量，本文提出一种应用危险品运输车辆轨迹数据的深度学习方法。针对车辆实时定位数据常有缺失，难以精准定位的问题，构建考虑多头自注意力机制的轨迹插补模型，以补全缺失轨迹；结合危险品运输特征，设计融合自适应机制和阈值规则的双层聚类算法识别车辆起讫点时间和经纬度信息；运单的精确起讫点名称可根据文本识别模型从聚类中心点周边兴趣点中筛选识别，从而实现对运单的增强。应用广东省液化天然气运输车数据测试所提出方法，结果表明：轨迹插补模型在不同缺失率下的平均绝对误差和平均绝对百分比误差值分别为2.34~3.33和6.05~7.74，均小于对比模型；危险品货运起讫点时间和位置识别的准确率为98.35%；文本识别模型对具体兴趣点的识别准确率达92.83%。本文构建方法可实现对预填运单关键字段的高精度反演与校正，从而对危险品运输安全管理提供重要的理论依据和实践指导。

关键词: 公路运输, 危险品运单增强, 起讫点识别, 轨迹插补, 双层聚类算法, 文本识别

Abstract: The waybills for the transportation of hazardous goods are usually filled in and reported by transport agencies before shipment in China. Therefore, the related time stamps of origin and destination are often inaccurate, which limit the effectiveness in transport management. A deep learning approach using the trajectory data of hazardous goods transportation vehicles is proposed to improve the quality of waybills. A trajectory imputation model incorporating a multi-head self-attention mechanism is firstly constructed to complete missing trajectories, which can solve the positioning accuracy problems caused by the missing real time vehicle positioning data. Then, based on the characteristics of hazardous goods transportation, a dual-layer clustering algorithm integrating an adaptive mechanism and multiple threshold rules is designed to identify the times and geographic coordinates of origin and destination for a specific waybill. The waybill can be enhanced by extracting precise origin and destination addresses from points of interest around the cluster centers using a text recognition model. The proposed methods were tested using the data of LNG transport vehicles in Guangdong, China. The results indicate that the Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) of the trajectory imputation model ranged from 2.34~3.33 and 6.05~7.74 under various data missing rates respectively, which outperform other baseline models. The accuracy of identifying the times and geographic coordinates of origin and destination reaches 98.35%. The text recognition model achieves an accuracy of 92.83% in identifying the address information of selected point of interests(POI). The proposed method enables high-precision inversion and correction of key fields in pre-filled waybills, thereby providing important theoretical support and practical guidance for the safety management of hazardous materials transportation.

Key words: highway transportation, enhancement of hazardous goods waybill, identification of origin and destination, trajectory imputation, dual-layer clustering algorithm, text recognition

中图分类号:

U491.6

陈邦举, 罗义凯, 陈磊磊, 伍翰廷, 项昌乐. 面向特征增强的危险品运单深度分析方法[J]. 交通运输系统工程与信息, 2025, 25(4): 306-316.

CHEN Bangju, LUO Yikai, CHEN Leilei, WU Hanting, XIANG Changle. A Deep Learning Approach for Enhancement of Hazardous Goods Waybill's Feature[J]. Journal of Transportation Systems Engineering and Information Technology, 2025, 25(4): 306-316.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2025.04.028

http://www.tseit.org.cn/CN/Y2025/V25/I4/306

参考文献

[1] SANCHEZ-DIAZ I. Assessing the magnitude of freight traffic generated by office deliveries[J]. Transportation Research Part A: Policy and Practice, 2020, 142: 279 289.

[2] PAPADOPOULOS A A, KORDONIS I, DESSOUKY M M, et al. Personalized Pareto-improving pricing-and routing schemes for near-optimum freight routing: An alternative approach to congestion pricing[J]. Transportation Research Part C: Emerging Technologies, 2021, 125: 103004.

[3] CHEN Y, YU P, CHEN W, et al. Embedding-based similarity computation for massive vehicle trajectory data [J]. IEEE Internet of Things Journal, 2021, 9(6): 4650 4660.

[4] LONG J A. Kinematic interpolation of movement data[J]. International Journal of Geographical Information Science, 2016, 30(5): 854-868.

[5] WINKEL R. On a generalization of bernstein polynomials and bezier curves based on umbral calculus (II): De casteljau algorithm[J]. Computer Aided Geometric Design, 2015, 39: 1-16.

[6] YUKSEL C, SCHAEFER S, KEYSER J. Parameterization and applications of Catmull-Rom curves[J]. Computer Aided Design, 2011, 43(7): 747-755.

[7]徐金华,罗义凯,李昱燃,等.基于时频分解与深度学习的轨道客流预测[J].山东大学学报(工学版),2024, 54(2): 60-68, 79. [XU J H, LUO Y K, LI Y R, et al. Rail passenger flow prediction based on time-frequency decomposition and deep learning[J]. Journal of Shandong University (Engineering Edition), 2024, 54(2): 60-68, 79.]

[8] MORITZ S, BARTZ-BEIELSTEIN T. ImputeTS: Time series missing value imputation in R[J]. Neural Computing and Applications, 2017, 9(1): 207-218.

[9] WANG J, WU N, LU X, et al. Deep trajectory recovery with fine-grained calibration using Kalman filter[J]. IEEE Transactions on Knowledge and Data Engineering, 2019, 33(3): 921-934.

[10] LARANJEIRO P F, MERCHÁN D, GODOY L A, et al. Using GPS data to explore speed patterns and temporal fluctuations in urban logistics: The case of São Paulo, Brazil[J]. Journal of Transport Geography, 2019, 76: 114 129.

[11] YANG Y, CAI J, YANG H, et al. TAD: A trajectory clustering algorithm based on spatial-temporal density analysis[J]. Expert Systems with Applications, 2020, 139: 112846.

[12] YANG Y, JIA B, YAN X Y, et al. Identifying intercity freight trip ends of heavy trucks from GPS data[J]. Transportation Research Part E: Logistics and Transportation Review, 2022, 157: 102590.

[13] BAI X, XIE Z, XU X, et al. An adaptive threshold fast DBSCAN algorithm with preserved trajectory feature POInts for vessel trajectory clustering[J]. Ocean Engineering, 2023, 280: 114930.

[14] 龙雨欣, 蒲亦非, 张卫华.基于分数阶高斯噪声的BERT 情感文本分类研究[J].四川大学学报(自然科学版), 2024, 61(4): 127-132. [LONG Y X, PU Y F, ZHANG W H, et al. A study on bert sentiment text classification based on fractional order gaussian noise[J]. Journal of Sichuan University (Natural Science Edition), 2024, 61(4): 127-132.]

[15] LI Y, WANG Q, ZHANG J, et al. The theoretical research of generative adversarial networks: An overview [J]. Neurocomputing, 2021, 435: 26-41.

[16] 徐东伟,彭航,商学天,等.基于图自编码-生成对抗网络的路网数据修复[J]. 交通运输系统工程与信息, 2021, 21(6): 33-41. [XU D W, PENG H, SHANG X T, et al. Road network data restoration based on graph self encoding-generative adversarial networks[J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(6): 33-41.]

[17] 王力, 李敏, 闫佳庆,等.基于生成式对抗网络的路网交通流数据补全方法[J].交通运输系统工程与信息, 2018, 18(6): 63-71. [WANG L, LI M, YAN J Q, et al. Traffic flow data complementation method for road network based on generative adversarial network[J]. Journal of Transportation Systems Engineering and Information Technology, 2018, 18(6): 63-71.]

[18] NI Q, CAO X. MBGAN: An improved generative adversarial network with multi-head self-attention and bidirectional RNN for time series imputation[J]. Engineering Applications of Artificial Intelligence, 2022, 115: 105232.

[19] 冯少荣, 肖文俊.基于密度的DBSCAN聚类算法的研究及应用[J]. 计算机工程与应用,2007(20): 216-221. [FENG S R, XIAO W J. Research and application of density-based DBSCAN clustering algorithm[J]. Computer Engineering and Applications, 2007(20): 216 221.]

[20] 黄钢, 瞿伟斌,许卉莹.基于改进密度聚类算法的交通事故地点聚类研究[J]. 交通运输系统工程与信息, 2020, 20(5): 169-176. [HUANG G, QU W B, XU H Y. Research on traffic accident location clustering based on improved density clustering algorithm[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(5): 169-176.]

[21] YANG Y, JIA B, YAN X Y, et al. Identifying intracity freight trip ends from heavy truck GPS trajectories[J]. Transportation Research Part C: Emerging Technologies, 2022, 136: 103564.

面向特征增强的危险品运单深度分析方法

A Deep Learning Approach for Enhancement of Hazardous Goods Waybill's Feature

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	徐进, 罗松, 陈礼彪, 李涛, 王涛. 高速公路应急车道开放工况车辆运行特性[J]. 交通运输系统工程与信息, 2025, 25(3): 372-382.
[2]	杜鹏. 新时期我国交通运输业碳减排的任务与对策[J]. 交通运输系统工程与信息, 2024, 24(6): 1-4.
[3]	李雷, 钱思达, 徐黎庆, 焦文玲, 张鑫, 张承虎. 基于生命周期评价的氢能重卡排放分析[J]. 交通运输系统工程与信息, 2024, 24(4): 292-299.
[4]	刘永红, 张帆, 苗领, 蔡雨锋, 赖雨梦, 吴潇彬 , 曾雪兰, 姚达文. 广东不同区域汽车新能源化的减碳潜力及成本预测[J]. 交通运输系统工程与信息, 2024, 24(1): 299-310.
[5]	郭红霞, 倪少权, 贺裕雁. 网络条件下供需未匹配的公路港多车型甩挂调度优化[J]. 交通运输系统工程与信息, 2023, 23(5): 290-297.
[6]	王直欢, 张文. 基于双层规划的氢能重卡车队替换研究[J]. 交通运输系统工程与信息, 2023, 23(4): 55-60.
[7]	马昌喜, 薛凡松, 麻存瑞, 李海军. 时变路网下基于混合调整策略的生鲜品配送路径优化研究[J]. 交通运输系统工程与信息, 2023, 23(4): 298-306.
[8]	谢如鹤, 何佳雯, 邹毅峰, 李庆庭, 刘子玲, 廖晶. 冷链零担物流前端集货至末端配送全链条优化模型[J]. 交通运输系统工程与信息, 2023, 23(3): 204-213.
[9]	王建伟, 刘旭旭, 付鑫, 杨洋, 崔梦妍. 一种“效率-公平-运力”多维权衡的需求可拆分配送方法[J]. 交通运输系统工程与信息, 2022, 22(5): 285-292.
[10]	王宁, 张佳蕊, 赵姣. 城际零担货运平台车辆路径问题研究[J]. 交通运输系统工程与信息, 2022, 22(2): 163-177.
[11]	谭二龙, 李宏海, 钟厚岳, 霍恩泽, 马晓磊. 基于轨迹数据的货车自发编队节油潜力估计[J]. 交通运输系统工程与信息, 2022, 22(1): 74-84.
[12]	贾术艳, 宋雨童, 杨紫都. 基于生长曲线函数的货车运营环节碳达峰研究[J]. 交通运输系统工程与信息, 2021, 21(6): 310-318.
[13]	甘蜜, 卿三东, 刘晓波, 李丹丹. 货车轨迹数据在公路货运系统中应用研究综述[J]. 交通运输系统工程与信息, 2021, 21(5): 91-101.
[14]	李根, 翟伟, 朱兴贝, 杨晟, 邬岚. 侧向碰撞风险对交织区汇合行为的影响[J]. 交通运输系统工程与信息, 2021, 21(4): 204-210.
[15]	李桃迎，王婷，张羽琪. 考虑多特征的高速公路交通流预测模型[J]. 交通运输系统工程与信息, 2021, 21(3): 101-111.